PREDIKSI LUAS PANEN DI KECAMATAN PURWOADADI MENGGUNAKAN ALGORITMA REGRESI LINEAR BERGANDA

Abstract

Agriculture, particularly rice cultivation, is highly vulnerable to climate change because it 
depends on water cycles and weather conditions to maintain productivity. Climate change 
affects crop growth, development, and yields, as agricultural activities are heavily dependent 
on weather and climate. This study utilizes data mining to introduce a new breakthrough in 
addressing rice farming issues in Grobogan Regency, Purwodadi District. The method used 
is multiple linear regression, with the dependent variable being harvested area and the 
independent variables including plxanted area and rainfall. The objective of this research is 
to test and develop data mining methods to predict yield levels, thereby assisting local 
governments in decision-making during crop failures, based on agricultural data from 2019
2023. The research process involves data collection, preprocessing, algorithm 
implementation, and result evaluation. The analysis shows that the multiple linear regression 
model provides reasonably accurate predictions, with a Root Mean Square Error (RMSE) 
value of 209.042 and a Relative Root Squared Error (RRSE) of 0.111. Furthermore, the 
analysis reveals that planted area significantly influence the harvested area. These findings 
offer insights for local governments as policymakers in providing aid during crop failures.